How ANSA picks winning AI niches

Marco DeMeireles, co founder and managing partner at ANSA, breaks down how a modern VC firm wins by being focused, data driven, and allergic to hype. If you want a clearer view of how investors evaluate open source, mission critical industries, and AI categories, this is a practical, operator minded look behind the curtain.

Marco explains ANSA’s focus on what they call undercover markets, from open source and open core businesses to defense, intelligence, cybersecurity, healthcare IT, and infrastructure companies that become deeply embedded and rarely lose customers. We also get into how they raised their first fund, why portfolio concentration changes everything, and how they push founders toward efficiency and profitability without killing ambition.

Key Takeaways

• In open source, two things matter more than most people admit: founder DNA tied to the project, and what you put behind the paywall that enterprises will pay for

• Concentration forces rigor, fewer bets means deeper diligence, clearer underwriting, and more hands on support post investment

• Great early stage support is not just advice, it is people, capital planning, and operating help that changes outcomes

• AI investing gets easier when you start with category selection, avoid fickle demand, then hunt for non obvious wedges in real workflows

• Long term winners tend to show compounding growth, improving efficiency, real demand, durable business models, founder strength, and an asymmetric risk reward at the price

Timestamped Highlights

00:00 Marco’s quick intro and what ANSA invests in

00:36 Undercover markets, open source, and mission critical industries explained

01:54 The two open source filters that change how ANSA underwrites a deal

03:31 Why open source can work in defense, plus the Defense Unicorns example

05:29 How a new firm raises a first fund, and what the right LP partners look for

10:50 The three levers ANSA pulls with founders: people, capital, operations

15:22 Marco’s six part framework for evaluating investments

17:39 How to tell who wins in crowded AI categories, and why niche wedges matter

21:41 The first investment they will never forget, and the air gapped cloud problem

A line worth stealing

“You can’t outsource greatness. You can’t outsource people selection.”

Pro Tips

• If you are building open source, be intentional about what is free versus paid, security, compliance, and auditability tend to earn real pricing power

• If your business depends on paid acquisition, test a path to organic growth early, it can unlock profitability and give you leverage in fundraising and exits

• In crowded AI spaces, pick a wedge where documentation is heavy, complexity is low, and ROI is obvious, then expand once you own that lane

Call to Action

If this episode helped you think more clearly about investing and building, follow the show, subscribe, and share it with one founder or operator who is navigating funding, pricing, or go to market right now